The Society for Immunotherapy of Cancer consensus ... › content › pdf › 10.1186 › s40425-016-0188-z.pdfDavid Avigan6, Lisa Barbarotta7, Austin John Barrett8, Koen Van Besien9,

POSITION ARTICLE AND GUIDELINES Open Access

The Society for Immunotherapy of Cancerconsensus statement on immunotherapyfor the treatment of hematologicmalignancies: multiple myeloma,lymphoma, and acute leukemiaMichael Boyiadzis1†, Michael R. Bishop2†, Rafat Abonour3, Kenneth C. Anderson4, Stephen M. Ansell5,David Avigan6, Lisa Barbarotta7, Austin John Barrett8, Koen Van Besien9, P. Leif Bergsagel10, Ivan Borrello11,Joshua Brody12, Jill Brufsky13, Mitchell Cairo14, Ajai Chari12, Adam Cohen15, Jorge Cortes16, Stephen J. Forman17,Jonathan W. Friedberg18, Ephraim J. Fuchs19, Steven D. Gore20, Sundar Jagannath12, Brad S. Kahl21, Justin Kline22,James N. Kochenderfer23, Larry W. Kwak24, Ronald Levy25, Marcos de Lima26, Mark R. Litzow27, Anuj Mahindra28,Jeffrey Miller29, Nikhil C. Munshi30, Robert Z. Orlowski31, John M. Pagel32, David L. Porter33, Stephen J. Russell5,Karl Schwartz34, Margaret A. Shipp35, David Siegel36, Richard M. Stone4, Martin S. Tallman37, John M. Timmerman38,Frits Van Rhee39, Edmund K. Waller40, Ann Welsh41, Michael Werner42, Peter H. Wiernik43

and Madhav V. Dhodapkar44*

Abstract

Increasing knowledge concerning the biology of hematologic malignancies as well as the role of the immune systemin the control of these diseases has led to the development and approval of immunotherapies that are resulting inimpressive clinical responses. Therefore, the Society for Immunotherapy of Cancer (SITC) convened a hematologicmalignancy Cancer Immunotherapy Guidelines panel consisting of physicians, nurses, patient advocates, and patientsto develop consensus recommendations for the clinical application of immunotherapy for patients with multiplemyeloma, lymphoma, and acute leukemia. These recommendations were developed following the previouslyestablished process based on the Institute of Medicine’s clinical practice guidelines. In doing so, a systematic literaturesearch was performed for high-impact studies from 2004 to 2014 and was supplemented with further literature asidentified by the panel. The consensus panel met in December of 2014 with the goal to generate consensusrecommendations for the clinical use of immunotherapy in patients with hematologic malignancies. During thismeeting, consensus panel voting along with discussion were used to rate and review the strength of the supportingevidence from the literature search. These consensus recommendations focus on issues related to patient selection,toxicity management, clinical endpoints, and the sequencing or combination of therapies. Overall, immunotherapy israpidly emerging as an effective therapeutic strategy for the management of hematologic malignances. Evidence-based consensus recommendations for its clinical application are provided and will be updated as the field evolves.

Keywords: Cancer immunotherapy, Hematologic malignancies, Acute leukemia, Lymphoma, Multiple myeloma,Immunotherapy

* Correspondence: [email protected]†Equal contributors44Department of Hematology & Immunobiology, Yale University, 333 CedarStreet, Box 208021, New Haven, CT 06510, USAFull list of author information is available at the end of the article

© The Author(s). 2016 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Boyiadzis et al. Journal for ImmunoTherapy of Cancer (2016) 4:90 DOI 10.1186/s40425-016-0188-z

http://crossmark.crossref.org/dialog/?doi=10.1186/s40425-016-0188-z&domain=pdfmailto:[email protected]://creativecommons.org/licenses/by/4.0/http://creativecommons.org/publicdomain/zero/1.0/

IntroductionThe incidence of hematologic malignancies has steadilyincreased over the past 30 years. Over this period oftime, there have been significant advancements in theunderstanding of the biology of these diseases, includingthe important role that the immune system plays in theirdevelopment, maintenance, and eradication. As a resultof these discoveries, there has been concurrent advance-ment in immunotherapies specifically developed for thetreatment of hematologic malignancies. Probably themost remarkable example of the success of immunother-apy for hematologic malignancies is the anti-CD20monoclonal antibody rituximab, which has been incor-porated into almost all aspects in the treatment of B cellmalignancies.An understanding of the basic mechanisms of the im-

mune system as it relates to hematologic malignancies hasbeen increasing rapidly. This understanding has acceler-ated the translation of this research and has led to thedevelopment of several novel immunotherapeutic ap-proaches. A major recent example is research related totumor immune evasion mechanisms. The programmedcell death-1 (PD-1) pathway has emerged as a highlyrelevant immune checkpoint pathway in a number ofhematologic malignancies, particularly Hodgkin’s lymph-oma [1]. This work has led to the development of severalantibodies that disrupt the interactions between negativeregulatory receptors on tumor-specific T cells and theirligands on tumor cells or antigen-presenting cells.In response to the growing number of immunothera-

peutic agents that have been approved and are in finalstages of clinical investigation in the treatment ofhematologic malignances, SITC formed a hematologicmalignancy Cancer Immunotherapy Guidelines panel toprovide guidance to practicing clinicians caring forpatients with multiple myeloma, lymphoma, and acuteleukemia. SITC is a nonprofit professional organizationdedicated to the basic understanding and clinical applica-tions of cancer immunotherapy. The panel consisted ofexperts in hematologic malignancies, including physicians,nurses, patient advocates, and patients (Additional file 1).This panel met to consider issues related to patient selec-tion, toxicity management, treatment cessation guidelinesand current recommendations for treatment sequencingwith the goal of preparing a consensus statement on clin-ical use of immunotherapy for patients with hematologicmalignancies. The hematologic malignancy panel wascomprised of three separate disease-specific panels fo-cused on multiple myeloma, lymphoma, and acuteleukemia (Fig. 1). The consensus panels were charged toprovide evidence-based guidelines and recommendationswith a major emphasis on US Food and Drug Administra-tion (FDA)-approved agents. While the members of thepanel agreed that allogeneic hematopoietic stem cell

transplantation (HSCT) is an important and effectivetherapeutic option in the management of hematologicmalignancies, it was not included in the current consensusstatement at the recommendation of the Steering Com-mittee. Although the major emphasis of this report is toprovide summaries and recommendations relative toapproved agents, the panel felt it was also important toaddresses biological principles and treatment that wouldbe relevant to clinical oncologists in regard to the futureof immunotherapy research for hematologic malignancies.

MethodsConsensus statement developmentThis consensus statement was developed using the stan-dards delineated by the SITC consensus statement ontumor immunotherapy for the treatment of cutaneousmelanoma as described previously [2]. These standardswere originally developed based on the Institute of Medi-cine’s Standards for Developing Trustworthy ClinicalPractice Guidelines, and include key components such asestablishing a transparent process for guideline develop-ment and funding, managing and reporting conflicts ofinterest, including a multidisciplinary and balanced panel,establishing an evidence-based foundation and ratingsystem for the strength of the evidence, reporting the re-sults through a publicly available website and publication,and having a plan to update the recommendations [2, 3].In December 2014, SITC convened a hematologic malig-

nancy Cancer Immunotherapy Guidelines panel chargedwith developing clinical practice guidelines for the use ofimmunotherapy in multiple myeloma, lymphoma, andacute leukemia. To do so, these Steering Committee-ledpanels considered patient selection, toxicity management,assessment of response, and sequencing as well as thecombination of therapies for immunotherapies in currentclinical practice. Due to differences in the regulation andavailability of immunotherapy agents world-wide, theconsensus panel focused on drugs currently approved bythe US FDA. These consensus guidelines are not intendedto be a substitute for the professional judgment of treatingphysicians. The full consensus recommendations as well asany future updates can be found on the SITC website [4].

Consensus panel and conflicts of interestPotential consensus panel members including physicians,nurses, patient advocates, and patients were solicited fromSITC members and non-members. Panel members werescreened using the SITC conflicts of interest disclosureform. This form requires the disclosure of any financial aswell as non-financial conflicts of interests that may havedirect implications resulting from the publication of thisstatement. In addition, no commercial funding was usedto support the consensus panel meeting, literature review,or preparation of this manuscript.

Boyiadzis et al. Journal for ImmunoTherapy of Cancer (2016) 4:90 Page 2 of 25

The hematologic malignancy panel, consisting of threeseparate disease-specific panels for multiple myeloma,lymphoma, and acute leukemia, met in December, 2014,to review and discuss results from a previously distrib-uted questionnaire collecting information on the panelmember’s role in patient care, primary clinical focus,experience with FDA-approved agents, and current clin-ical practices concerning the use of, or recommendeduse of immunotherapy agents. The final version of thisconsensus statement was made available to the entireSITC membership for an open comment period. Thesecomments were collected and considered in the finalversion of this manuscript (Additional file 2).

Literature reviewThe MEDLINE database was used to perform a system-atic search of scientific literature from 2004 to 2014.The search was limited to “humans” and “clinical trialsor controlled clinical trials or randomized controlledclinical trials.” The results from the literature search arelisted according to each disease type as follows. Thesebibliographies were supplemented with additional litera-ture as identified by the panel.

Multiple myelomaThe search terms included “myeloma and lenalidomide,”“myeloma and pomalidomide,” “myeloma and thalido-mide,” “myeloma and monoclonal antibody,” “myelomaand checkpoint blockade or PD-1 or PD-L1 or B7-H1,”“myeloma and oncolytic virus,” “myeloma and virother-apy,” and “myeloma and dendritic cell vaccine or idiotypevaccine.” After duplicates and irrelevant citations were

removed, this search resulted in a 173-item bibliography(Additional file 3: Bibliography I).

LymphomaThe search terms included “lymphoma and rituximab orofatumumab,” “lymphoma and checkpoint blockade,”“lymphoma and chimeric antigen receptor,” “lymphomaand idiotype vaccine,” “lymphoma and denileukin diftitox,”“lymphoma and interferon alfa-2b,” “mantle cell lymphomaand lenalidomide,” and “mantle cell lymphoma and borte-zomib.” After duplicates and irrelevant citations wereremoved, this search resulted in a 138-item bibliography(Additional file 3: Bibliography II).

Acute leukemiaThe search terms included “AML and epigenetic therapy,”“AML and hypomethylating agents or 5-azacytidine ordecitabine,” “AML and monoclonal antibody,” “ALL andmonoclonal antibody or rituximab or blinatumomab,”“AML and checkpoint blockade,” “AML and CAR orCART,” and “ALL and CAR or CART.” After dupli-cates and irrelevant citations were removed, thissearch resulted in a 56-item bibliography (Additionalfile 3: Bibliography III).The literature was reviewed and graded according to

the previously established rating system [2]. In summary,Level A was defined as strong supporting evidence-based data from prospective, randomized clinical trials,and meta-analyses; Level B was defined as moderatesupporting data from uncontrolled, prospective clinicaltrials; and Level C represented weak supporting datafrom retrospective reviews and case reports.

Fig. 1 Table of the Cancer Immunotherapy Guidelines for Hematologic Malignancy participants. Asterisks (*) indicate panel chair and steeringcommittee member


Multiple myelomaImmune-based therapies in multiple myeloma (MM) canbe classified as current or emerging therapies, basedlargely on the level of clinical evidence. The panel there-fore first considered the status of current therapies,followed by considerations for the current status andoptimal evaluation of emerging therapies.

Current immunotherapies in myelomaTwo broad categories of current immune/immune-modulating therapies in MM are immune-modulatingdrugs (IMiDs) and anti-tumor monoclonal antibodies(mAbs). Thalidomide, lenalidomide, and pomalidomideare already FDA-approved for use in MM [5, 6]. Whilenon-immune effects of IMiDs are recognized, the mye-loma panel voted to include these agents among the listof immune therapies for these guidelines. Although anti-tumor antibodies were not yet FDA-approved at thetime of the panel review, the level of evidence support-ing the clinical activity of some agents (anti-CD38 mAb(daratumumab) and anti-SLAMF7 mAb (elotuzumab))was felt to be high, and therefore, they were includedamong current immune therapies [7, 8]. Both elotuzu-mab and daratumumab recently received FDA approvalfor relapsed myeloma.

IMiDs: thalidomide, lenalidomide, and pomalidomideOver the past 15 years, the use of IMiDs together withproteasome inhibitors has transformed the therapeuticlandscape and outcome of patients with MM. Lenalido-mide plus dexamethasone (Rd) was superior to dexa-methasone alone in two phase III trials involvingpatients with relapsed/refractory MM (RRMM) [9, 10].Rd was also superior to dexamethasone in the setting ofinduction therapy [11]. Use of a lower dose of dexa-methasone led to an improved safety profile, and accord-ingly, Rd has been commonly adopted in the US [12]. Ina clinical trial involving elderly patients with previouslyuntreated MM, continuous Rd was superior to fixedduration Rd and to melphalan, prednisolone, andthalidomide (MPT) [13].The Rd regimen has also been combined with sev-

eral agents, most notably proteasome inhibitors. Datacomparing the addition of carfilzomib to Rd (KRd) inRRMM demonstrated improved progression-free sur-vival (PFS) [14]. In a phase III trial, the addition ofelotuzumab to Rd led to improved PFS in patientswith RRMM [15]. Recently, the addition of ixazomibto the Rd backbone also led to improved PFS inRRMM [16]. It should be noted that these phase IIIstudies were performed in patients with lenalidomide-sensitive disease, though differences in patient popula-tions preclude across study comparisons.

In the front-line setting, results from trials comparingRd to triplets such as those in combination with borte-zomib (VRd), carfilzomib, and elotuzumab are currentlyawaited. Initial data from SWOG 0777 have demon-strated superiority of VRd over Rd in front-line therapyof myeloma [17]. Data from randomized clinical trialsevaluating the timing of stem cell transplantation in theera of novel agents is also awaited. Initial data from aphase III trial demonstrated improvement in PFS inpatients receiving early stem cell transplantation [18].Lenalidomide has also been utilized in the setting ofmaintenance therapy following autologous HSCT asdemonstrated in clinical trials Cancer and LeukemiaGroup B (CALGB) 100104 and IFM 2005–02 or ascontinuous therapy for transplant ineligible patients(MM-015) [19–21]. All three trials reported significantdifferences in PFS, and the CALGB trial reported im-proved 3 year overall survival (OS).Pomalidomide plus dexamethasone demonstrated re-

markable activity in patients with RRMM refractory tolenalidomide, and was the last immunotherapy agentapproved for therapy of MM [22–24]. Two dosingschedules (2 mg daily or 4 mg on 21/28 day schedule) ofpomalidomide (in combination with dexamethasone)have been explored with comparable results [25–27].Pomalidomide is also active in patients with high-riskcytogenetics such as deletion 17 [28].In recent years, the E3 ubiquitin ligase cereblon has

been identified as a key target of IMiDs [29, 30]. Bindingof the drug to cereblon leads to degradation of Ikarosfamily zinc finger proteins IKZF1 and IKZF3, which thenleads to inhibition of tumor cell growth and immuneactivation [31–33]. In preclinical and early clinical stud-ies, immune activation by IMiDs provides the basis forsynergy in combination with vaccines, antibodies andcheckpoint inhibitors [34–37]. IMiD therapy leads toactivation of both T and natural killer (NK) cells in vivo[27, 38, 39]. IMiD-mediated immune activation is rapidand correlates with clinical response to therapy [27].Myeloma Panel Recommendations:

� The panel recommends the use of combinationtherapies with lenalidomide in both front-line andrelapsed MM setting based on level A evidence.Data directly comparing regimens commonlyutilized in front-line setting is awaited andenrollment in well-designed clinical trials isrecommended. In a recent SouthwesternOncology Group (SWOG) study, combinationtherapy with VRd led to improved outcomecompared to Rd [17].

� The front-line regimen for transplant-eligiblepatients (outside of a clinical trial) preferred by themajority (53.3%) of the panel was VRd, followed by


Rd (26.7%) and cyclophosphamide, bortezomib, anddexamethasone (CyBorD) (13.3%) based on level Bevidence.

� The front-line regimen for transplant-ineligiblepatients (outside of a clinical trial) preferred by thepanel were Rd (46.6%), VRd (40%), followed byCyBorD (6.7%) based on level B evidence.

� Based on the results of SWOG S0777 (not availableat the time of the panel review), VRd is nowexpected to become the preferred front-line regimenfor most patients with newly diagnosed MM basedon level A evidence. Participation in ongoing clinicaltrials comparing this regimen with others is stronglyencouraged.

� All panelists recommend the use of a proteasomeinhibitor-based regimen in patients with t(4:14), del17p,and plasma cell leukemia based on level B evidence.

� The panel recognizes the lack of level A evidenceregarding the timing of stem cell transplant in theera of novel agents. While the results of studiesaddressing these questions are awaited, mostpanelists (66.7%) favor consideration of earlyautologous HSCT. The outcome of the Frenchcohort Intergroupe Francophone Du Myeloma trialhas been recently presented and demonstratedimproved PFS with early transplant. These data werenot available at the time of the panel review [18].

� The majority of the panel (80%) recommends theuse of maintenance therapy following autologousHSCT based on level A evidence. The preferredduration of maintenance therapy is until progression(50% of panelists) or for 2 years (28.6% of panelists).Patients on lenalidomide maintenance after priormelphalan exposure should also be monitored forsecondary malignancies.

� Preclinical and clinical data support the design ofclinical studies combining IMiDs with severalimmune therapies including monoclonal antibodies,vaccines, and immune checkpoint inhibitors basedon level B evidence.

� Nearly all of the clinical data with IMiDs is incombination with concurrent steroids, includingthat in the setting of current combinations withmonoclonal antibodies. Although steroids have thepotential to dampen immune activation, recent datasuggests that IMiDs may be able to activate immunityeven in the setting of concurrent steroids [27, 40].The impact of concurrent steroids on IMiD-basedimmune therapies was debated, and the panel agreedthat minimizing (or eventually eliminating) steroidswould be highly desirable. However, there is a lack ofconsensus and currently no data to support the needto eliminate steroids, particularly in light of theirsynergistic direct anti-tumor effects.

Anti-tumor monoclonal antibodiesIn recent years, several anti-tumor mAbs have enteredclinical testing in MM. Of these, elotuzumab and daratu-mumab have entered phase III testing. Elotuzumab is afully humanized mAb against the glycoprotein SLAMF-7expressed on myeloma and NK cells [41]. In preclinicalmodels, elotuzumab illustrated anti-tumor effects via NKactivation and enhanced antibody-dependent cytotoxicity[41]. In a phase II trial, elotuzumab plus Rd (Elo-Rd)achieved a 92% objective response rate (ORR) in patientswith RRMM [42]. In a recent phase III trial, Elo-Rd led toan improvement in PFS compared to Rd in patients withRRMM, including those with high-risk features [15]. Inthis study, median PFS was 19.4 months in the Elo-Rdgroup vs. 14.9 months in the Rd group alone, with ahazard ration of .70 (95% CI: .57-.85, P < .001).Daratumumab targets CD38 expressed on MM cells as

well as hematopoietic progenitor cells, endothelial cells,and activated immune cells [43]. Anti-myeloma effectsof daratumumab involve several mechanisms includingdirect as well as immune-mediated effects [44]. Prelim-inary studies with daratumumab showed promising sin-gle agent activity with 31% objective responses in heavilypretreated RRMM, including those refractory to bothproteasome inhibitors and IMiDs [45]. These resultswere confirmed in a phase I-II study, illustrating a 36%response rate and median PFS of 5.6 months in heavilypretreated RRMM patients who received daratumumabmonotherapy (16 mg/kg) [46]. In addition, in a phaseII, multicenter trial daratumumab showed a 29.2% re-sponse rate and median PFS of 3.7 months in MMRRpatients who had received a median of 5 previous linesof therapy [47]. Moreover, the addition of daratumu-mab to the Rd backbone led to an improved ORR of75% in RRMM. Daratumumab has also been combinedwith pomalidomide in therapy of patients with RRMM[48]. Similar results have been observed with anotheranti-CD38 mAb, SAR650984 (isatuximab) in patientswith RRMM.Two antibody-drug conjugates (ADCs) are in active

clinical testing in RRMM. Indatuximab ravtansine(BT062) is comprised of an anti-CD138 mAb conju-gated to the maytansinoid DM4 toxin. In a phase IItrial, indatuximab ravtansine plus Rd led to a 78%ORR in patients with RRMM. J6MO-mcMMAF(GSK2857916) is an ADC targeting B cell maturationantigen currently in phase I testing in RRMM. Inaddition, mAbs targeting several other molecules (e.g.,CD40, CD56, CD54) are also in preclinical/early clin-ical testing. mAbs may be of particular interest inpopulations at higher risk with current therapies, in-cluding those with genetic high risk disease andcomorbidities such as renal failure.Myeloma Panel Recommendations:


� mAbs targeting SLAMF-7 (elotuzumab) or CD38(daratumumab and SAR650984) in combinationwith Rd or VRd have demonstrated promisingclinical activity in RRMM, including those withhigh-risk disease. Eligible patients with RRMM orNDMM and particularly those with high-riskfeatures should be encouraged to participate inongoing clinical trials with these agents based onlevel B evidence. After the panel meeting, onNovember 16, 2015, daratumumab receivedapproval to treat patients with relapsed MM whohave received at least three prior lines of therapy orare refractory to both a proteasome inhibitor and anIMiD. On November 30, 2015, the FDA approvedelotuzumab in combination with lenalidomide anddexamethasone for therapy of relapsed MM whohave received one to three prior medications.

� IMiDs often show synergy with mAbs likely in partrelated to their effects on antibody-dependent cellmediated cytotoxicity (ADCC) and are emerging asimportant agents for combination with mAbs,although proteasome inhibitors are also beingcombined with monoclonal antibodies.

Emerging immunotherapies in myelomaFor the evaluation of emerging therapies, the panel con-sidered both early phase clinical as well as key preclinicalfindings from the literature in its recommendations. It isrecognized that this is an area of active ongoing preclinicaland clinical investigation with several new approachesshowing promise. Therefore, periodic updates to theserecommendations are strongly recommended.

Immune checkpoint blockadeSeveral studies have shown that PD-L1 is commonlyoverexpressed by myeloma tumor cells [49]. In preclin-ical models, targeting PD-L1 led to anti-tumor effects inmurine myeloma [50]. Blockade of the PD-L1 axis leadsto activation of antigen-specific T and NK cells inculture [36, 51, 52]. Expression of PD-L1 in MM tumorcells is enriched in minimal residual disease and correlateswith risk of progression from monoclonal gammopathy ofundetermined significance (MGUS) to MM [53, 54]. Inphase II clinical studies with the anti-PD-1 antibody nivo-lumab, stable disease (but no objective regressions) wereobserved in RRMM patients [55]. The impact of targetingthis axis on survival of MM patients is currently unknown.Early data combining anti-PD-1 antibody pembrolizumabwith IMiDs (lenalidomide and pomalidomide) have beenreported and suggest promising clinical activity. Limitedsingle agent activity with PD-1 blockade in early myelomastudies suggests the need to consider combination withother agents or approaches that stimulate and expandtumor specific lymphocytes [56, 57].

Myeloma Panel Recommendations:

� There was a consensus among the panel for a strongpreclinical rationale for consideration of clinicaltrials of immune-checkpoint blockade in myeloma.

� The panel identified the following top clinical settingsfor evaluation of immune checkpoint blockade assingle agents: high-risk MM, post-autologous HSCT,and minimal residual disease (MRD).

� The panel identified the following top clinicalsettings for evaluation of immune checkpoint-basedcombination therapies: relapsed MM, high-risk MM,and post-autologous HSCT.

� The panel identified the following as the top threeagents for combination with immune checkpointblockade in clinical trials: lenalidomide/IMiDs,vaccine, and other immune checkpoint inhibitors.Update added after the panel meeting: initial reportsof studies testing combination of IMiDs and immunecheckpoint blockade have shown promising clinicalactivity. Tumor-directed mAbs are also attractiveagents for combination with immune checkpointblockade. Thus, participation in phase II/III trialstesting these combinations is strongly encouraged.

Immune activating antibodiesThere are preclinical data to support targeting co-stimulation via activating antibodies in MM. Oneexample is targeting CD137, which leads to antitumoreffects in mouse models [58, 59]. Targeting CD137 hasalso been shown to synergize with anti-tumor antibodiesin preclinical models [60–62].Myeloma Panel Recommendations:

� There is preclinical rationale to consider clinicalevaluation of immune activating antibodies in MM.

� The panel identified the following top clinicalsettings for evaluation of immune activatingantibodies as single agents: relapsed MM, MRD, andpost-autologous HSCT.

� The panel identified the following top clinicalsettings for evaluation of immune activatingantibody-based combination therapies: high-riskMM, MRD, and post-autologous HSCT.

� The panel identified the following as the top agentsfor combination with immune-activating antibodiesin clinical trials: lenalidomide/IMiDs and vaccines.With the emergence of anti-tumor antibodies,there is interest in combining these with immuneactivating antibodies as well.

VaccinesVaccines against tumor-specific antigens represent an at-tractive strategy to boost tumor immunity and may be


particularly relevant with the emergence of checkpointblockade strategies. Most of the early vaccine studies inMM targeted idiotypic determinants on clonal immuno-globulin (Ig) [63–65]. Ongoing vaccine studies aretargeting peptides derived from defined antigens, incombination with lenalidomide and with anti-PD-1 [66].Several vaccine approaches are in early phase testing.The PVX-410 vaccine consists of a cocktail of HLA-A2derived peptides from X-box binding protein1 (XBP-1),CD138, and SLAM-F7 antigens that can trigger activa-tion of MM-specific T cells and is currently under evalu-ation in combination with lenalidomide and anti-PD-1(NCT01718899). One particular approach to boost im-munity to multiple tumor-associated antigens involvesfusion of tumor cells and dendritic cells (DCs) [67–69].In a phase II trial, MM-DC vaccination following autolo-gous HSCT led to a 78% very good partial response(VGPR) rate, and a 47% complete response (CR)/nearcomplete response (nCR) rate, with responses improvingfrom PR to CR/nCR after 100 days in 24% of patients[70]. This approach is now being tested in a randomizedmulticenter clinical trial. DC vaccines targeting innatelymphocytes such as NKT cells in combination withlow-dose lenalidomide also led to tumor regression inasymptomatic MM in a small clinical trial [71]. Anotherapproach has been to use an allogeneic myeloma vac-cine in combination with a GM-CSF secreting cell line(myeloma GVAX) [72]. When administered in combin-ation with lenalidomide in patients in a near completeremission (with a detectable immunofixation of theirmonoclonal protein), patients have shown evidence ofpriming and persistence of a tumor-specific immune re-sponse that correlated with an ongoing disease remission[73]. These data have led to a randomized trial comparinglenalidomide maintenance to lenalidomide +GVAX.Myeloma Panel Recommendations:

� Vaccines represent an attractive strategy to boosttumor-specific immunity, particularly in the settingof early phase or MRD [70, 71, 74].

� The panel identified MRD and high-risk asymptom-atic MM as the top clinical settings for clinicalevaluation of vaccine strategies.

� Clinical evaluation of vaccines is stronglyrecommended in combination with approaches thatmodify the immune suppressive factors in the tumormicroenvironment. The panel identifiedlenalidomide and immune checkpoint blockade asthe top strategies for combination with vaccines.

Adoptive cellular therapies, including chimeric antigenreceptor (CAR) T cellsAdoptive transfer of activated tumor-infiltrating T cellsled to tumor regression in patients with melanoma. In a

similar fashion, marrow infiltrating T cells have beeninfused following ex vivo activation in MM patients fol-lowing autologous HSCT. In a recent study with 25patients treated using this approach, the presence ofcentral memory a CD8+ T cell phenotype at baselineand persistence of myeloma-specific T cells at 1 yearpost adoptive T cell therapy was predictive of improvedoutcome [75, 76]. One strategy involved combiningvaccination against tumor antigens with adoptive trans-fer of anti-CD3-stimulated and vaccine-primed T cellsfollowing autologous HSCT in patients with RRMM[77–79]. Antigens targeted via this approach included h-TERT and survivin in one study and MAGE in anotherstudy [77, 78]. The combination approach led to en-hanced reconstitution of cellular and humoral immunitypost-ASCT, including tumor-specific T cells.CAR T cells against CD19 have shown remarkable clin-

ical activity in acute lymphoblastic leukemia (ALL) [80].CART-19 cells are currently being evaluated in the settingof MM following autologous HSCT, based on the premisethat a subset of drug resistant and possibly clonogenicsubset of tumor cells express CD19 [81] and have shownearly signs of activity [81]. Another antigen being targetedin early phase clinical trials by this approach is B cell mat-uration antigen [82], and NY-ESO-1 has been targetedwith TCR-engineered T-cells [83]. Other approaches test-ing CAR-modified T or NK cells are targeting diverseantigens such as kappa light chain, NKG2D, CD38 andSLAMF-7. In addition to cell-based therapies, virotherapyapproaches such as measles virus have also been evaluatedin patients with RRMM and impressive clinical responseshave been observed in some patients with this approach[84]. Virus-induced death of tumor cells is thought to acti-vate anti-tumor immunity, which sets the stage forcombination approaches [85].Myeloma Panel Recommendations:

� Adoptive transfer of costimulated/vaccine-primed Tcells as well as marrow infiltrating T cells is apromising strategy for immunotherapy of MM.

� Several CAR-modified T/NK cell approaches arealso being developed and in preclinical/early phasetesting.

� Virotherapy approaches such as measles virus haveled to impressive clinical responses in some patientswith RRMM.

� The panel identified patients with high-risk MM orRRMM as well as post-autologous HSCT as pre-ferred clinical settings for clinical evaluation ofadoptive cellular therapies.

� The panel also identified combination approacheswith lenalidomide and immune checkpoint blockadeas preferred combination approaches with thesestrategies.


Issues related to immunotherapy research in myelomaEmergence of effective immune therapies in cancer hasled to a reassessment of trial designs and endpoints forevaluating clinical efficacy of such therapies, particularlyin the setting of some solid tumors. Traditional criteriasuch as response rates and PFS did not correlate withOS or clinical benefit for some immune therapies in thesetting of solid tumors. Novel immune related responsecriteria have been proposed in the setting of some solidtumors [86].Prior preclinical studies have shown that tumor-specific

T cells are enriched in the bone marrow in preneoplasticgammopathy and even in the setting of clinical MM, Tcells from the bone marrow can be activated to kill autolo-gous tumor cells [76, 87, 88]. Antigen-specific T cells havebeen detected in both blood and bone marrow of mye-loma patients [89, 90]. The phenotypic and functional pro-file of immune cells in the bone marrow differs from thatin circulation, such as with accumulation of IL17-producing T cells [91–94]. MM patients may have signifi-cant immune paresis in terms of both humoral and cellu-lar immunity, which may also be impacted by priortherapies [95]. Detection of MRD is emerging as an im-portant parameter and further research is needed to fullyintegrate MRD testing in the management of myeloma.Myeloma Panel Recommendations:

� The panel strongly recommends incorporation ofdetailed immune monitoring in ongoing clinical trialsof immune therapies including IMiDs, mAbs and otheremerging immune therapies based on level A evidence.

� The panel recommends that immune monitoringshould include serial analysis of the bone marrowmicroenvironment in all studies, as this may differfrom the findings in circulating immune cells basedon level A evidence.

� Immune monitoring should include both phenotypicand also functional studies including analyses ofantigen-specific T cell responses. Guidelines foroptimal monitoring of tissue-based immuneresponses, including those in the bone marroware currently under development through SITC.Collection, initial processing, transport andstorage of tissue aspirates or biopsies may have animpact on results of immune monitoring approaches,and these details should be included in the clinicalprotocols as well as publication of results.

� Timing of immune monitoring may depend on thenature of the specific therapy. For example, mid-cyclemeasurements may be needed to fully evaluate theeffect of IMiDs [71].

� The nature of preexisting immune paresis mayimpact the response to immune therapies andshould be considered in trial design [95].

� The panel concluded that there are insufficient datato evaluate whether current criteria for clinicalresponse/progression are inadequate for theevaluation of response to immune therapies andwhether immune related response criteria as in thesetting of solid tumors will be useful in MM.Nonetheless, repeat tumor biopsies should bestrongly considered to confirm disease progressionand avoid the potential caveat of pseudoprogressiondue to a transient increase in M protein or thepossibility that progression by imaging may reflectimmune infiltration as opposed to true progression.

� The panel concluded that there were insufficientdata at present to recommend a change in preferredendpoints for MM clinical trials in immunotherapy.However, the panel did note that PFS has not been aconsistent or reliable predictor of eventualimprovement in OS following immune therapies insolid tumors. It is possible that PFS at a defined time-point (e.g., 2 or 3 years) may be a better correlate ofclinical benefit with immune therapies, but this hasnot been validated.

LymphomaThe overall goal of the lymphoma consensus panel was toprovide guidance on the use of immunotherapeutics topracticing physicians caring for patients with lymphoma.The specific goal was to provide evidence-based guidelinesand recommendations with a major emphasis on FDA-approved agents. In addition, the panel was charged toprovide consensus opinions relative to: 1) defining optimalselection of lymphoma patients for immunotherapy; 2)improving management of immunotherapy side effects; 3)how best to monitor responses to immunotherapy; and 4)developing a rationale for sequencing (or combining) im-munotherapy with other agents for patients with high-riskand advanced disease.

Definition of an immunotherapeutic agentFor the purpose of their review, the panel initially ad-dressed how to define whether an agent or therapy wasa form of immunotherapy. In the broad sense, severaltherapeutic agents may have effects upon the immunesystem, but it may not be their major mechanism ofaction in the treatment of lymphoma. It was the consen-sus opinion that the major mechanism of action of alymphoma immunotherapeutic agent was augmentinganti-tumor responses of immune cells. For example, ifan agent directly inhibits tumor escape mechanisms, itwould be classified as immunotherapy. In contrast,agents that target a tumor cell directly and mediate celldeath mostly through non-immunological pathways (e.g.,targeted agents to B cell receptor) were not consideredimmunotherapeutics. Based on this definition, the list of


FDA-approved agents which the panel did not consideras a “true” form of immunotherapy for lymphoma in-cluded bortezomib, denileukin diftitox, brentuximabvedotin, temsirolimus and the radio-immunoconjugatesY-90 ibritumomab tiuxetan as well as tositumomab andiodine I-131 tositumomab.It was thoroughly recognized by the lymphoma panel

that allogeneic HSCT is an important and efficaciousform of immunotherapy in the treatment of lymphoma[96]. However, it was the recommendation of the steer-ing committee overseeing the hematologic malignanciespanels to not include this topic in the first set of guide-lines. It is the intent to review in a future update how toincorporate new immunotherapies into both allogeneicand autologous HSCT and how these agents may chal-lenge standard uses of allogeneic transplant.

Current immunotherapies in lymphomaMonoclonal antibodiesRituximabRituximab is a chimeric anti-CD20 mAb and is the mostcommonly used and most clearly defined immunotherapyin lymphoma. Rituximab is FDA-approved for the treat-ment of non-Hodgkin lymphoma (NHL) and chroniclymphocytic leukemia (CLL). Specifically, rituximab isindicated for the treatment of NHL patients with: 1) re-lapsed or refractory, low-grade or follicular, CD20-positive, B cell NHL as a single agent; 2) previouslyuntreated follicular, CD20-positive, B cell NHL in combin-ation with cyclophosphamide, vincristine, and prednisone(CVP) chemotherapy; 3) non-progressing (including stabledisease), low-grade, CD20-positive, B cell NHL, as a singleagent, after first-line CVP chemotherapy; and 4) previ-ously untreated diffuse large B cell, CD20-positive NHL incombination with cyclophosphamide, adriamycin, vincris-tine, prednisone (CHOP) or other anthracycline-basedchemotherapy regimens. Rituximab is also indicated, incombination with fludarabine and cyclophosphamide, forthe treatment of patients with previously untreated andpreviously treated CD20-positive CLL. Although it is wellrecognized that rituximab may have several mechanismsof action, the primary effect is on normal anti-tumor im-mune response [97]. It has been demonstrated that theFab domain of rituximab binds to the CD20 antigen onlymphocytes, and the Fc domain recruits immune effectorfunctions to mediate B cell lysis. Mechanisms of actioninclude direct anti-proliferative effects, complement-dependent cytotoxicity (CDC), and ADCC, with the latterbelieved to be dominant in vivo [98].Lymphoma Panel Recommendations:

� Rituximab is FDA-approved as maintenance therapyfor previously untreated follicular, CD20-positive Bcell NHL and in non-progressing, low-grade, CD20-

positive, B cell NHL after first-line CVP chemotherapy.However, the clinical benefit of maintenance rituximabin these two clinical settings remains controversial,based on endpoints that fail to clearly demonstrate asurvival benefit. It was the consensus opinion based onlevel B evidence that maintenance rituximab is not rec-ommended in low burden (as generally definedGroupe D’Etude de Lymphomes Folliculaires),low-grade NHL, and patients should be carefullycounseled relative to clinical benefits based onspecific endpoints [99, 100].

� Maintenance rituximab is not recommended indiffuse large B cell lymphoma (DLBCL) based onlevel A evidence.

� The panel further emphasized there are severalunresolved issues with endpoints used to assess theclinical utility of maintenance rituximab, as selectedendpoints may have differing relevance in differenthistologies (e.g., mantle cell lymphoma). Futuretrials addressing the role of maintenance rituximabshould clearly define and emphasize endpoints basedon histology.

� The panel could not make any recommendationsrelative to dose, frequency, and duration ofrituximab as maintenance therapy.

OfatumumabOfatumumab is a fully human anti-CD20 antibody thatis FDA-approved in combination with chlorambucil, forthe treatment of previously untreated patients with CLLfor whom fludarabine-based therapy is considered in-appropriate. The approval was based on the results of amulticenter randomized open-label trial that demon-strated improved PFS with ofatumumab in combinationwith chlorambucil as compared to single-agent chloram-bucil [101].Lymphoma Panel Recommendations:

� The panel had no specific recommendation forofatumumab as the results were not viewed asproviding any significant clinical advantages overrituximab. Ofatumumab is currently approved incombination with chlorambucil for front-linetherapy of CLL.

ObinutuzumabObinutuzumab is a humanized, glyco-engineered type 2,anti-CD20 antibody that is FDA-approved for use incombination with chlorambucil for the treatment ofpatients with previously untreated CLL. The approvalwas based on demonstration of an improvement in PFSin a randomized, open-label, multicenter trial comparingobinutuzumab in combination with chlorambucil tochlorambucil alone in patients with previously untreated


CD20-positive CLL. The study also included a rituximabin combination with chlorambucil arm [102].Lymphoma Panel Recommendations:

� The panel had no specific recommendation forobinutuzumab for lymphoma as the results in thisdisease, as opposed to CLL, were not viewed asproviding any significant clinical advantages overrituximab.

AlemtuzumabAlemtuzumab is a recombinant DNA-derived human-ized IgG1 kappa anti-CD52 monoclonal antibody indi-cated as a single agent for the treatment of B cell CLL.Alemtuzumab was initially FDA-approved in 2001 underaccelerated approval and subsequently to regular ap-proval based on an international, multicenter trial in 297previously untreated CLL patients randomized to eitheralemtuzumab or chlorambucil [103]. The PFS was sig-nificantly longer in the alemtuzumab arm; no differencesin survival were observed.Lymphoma Panel Recommendations:

� Alemtuzumab significantly impairs most importantimmunologic effectors and potentially impairs theutility of other immunotherapeutics.

� CD52 is expressed by approximately half of allperipheral T cell lymphomas, and alemtuzumab hasbeen used alone and in combination withconventional chemotherapy in their treatment.However, as with CLL, there is significant concernover toxicity and immunosuppression.

Other lymphoma immunotherapiesLenalidomideLenalidomide, a thalidomide analogue, is an immuno-modulatory agent with antiangiogenic and antineoplasticproperties. Lenalidomide is FDA-approved for the treat-ment of mantle cell lymphoma (MCL) that has relapsedor progressed after two prior therapies, one of whichincluded bortezomib. The approval of lenalidomide forMCL was based on a multicenter, single-arm, open-labeltrial of single-agent lenalidomide in 134 patients whoseMCL had relapsed after or were refractory to bortezo-mib or a bortezomib-containing regimen [104]. Treat-ment with lenalidomide resulted in an ORR of 26%; themedian duration of response was 16.6 months. Thecombination of lenalidomide plus rituximab (LR) hasbeen investigated as initial therapy in MCL [105]. In asingle-group, multicenter, phase 2 study, 38 patients withuntreated MCL received lenalidomide (20 mg/day x 21days of a 28-day cycle) as induction therapy for 12cycles. Rituximab was administered once weekly for thefirst 4 weeks and then once every other cycle until

disease progression. The most common grade 3 or 4adverse events were neutropenia (50%), rash (29%),thrombocytopenia (13%), an inflammatory syndrome (11%), anemia (in 11%), serum sickness (in 8%), andfatigue (in 8%). At the median follow-up of 30 months,the overall response rate in evaluable patients was 92%,and the CR rate was 64%. Median PFS had not beenreached at the time of this report. The 2-year PFS andOS were estimated to be 85% and 97%, respectively. Aresponse to treatment was associated with improvementin quality of life [105].In a multicenter phase II/III study, DLBCL patients

were stratified by germinal center B cell-like (GCB)versus non-GCB subtype, then randomized 1:1 to re-ceive lenalidomide or investigator’s choice (IC) chemo-therapy until progressive disease, unacceptable toxicity,or voluntary withdrawal [106]. Patients with GCB or non-GCB DLBCL treated with lenalidomide had similar ORR,but the data suggested greater improvements in PFS andOS with lenalidomide versus IC in the non-GCB patients,particularly the ABC subtype. In the Alliance phase II trial,patients with relapsed follicular lymphoma (FL) were ran-domized to rituximab alone or lenalidomide alone or LR[107]. The rituximab-alone arm was discontinued as aresult of poor accrual. The ORR was 53% (CR = 20%) and76% (CR = 39%) for lenalidomide alone and LR, respect-ively (P = 0.029). Patients were treated until time ofprogression. At the median follow-up of 2.5 years, mediantime to progression was 1.1 years for lenalidomide aloneand 2 years for LR (P = 0.0023).Lymphoma Panel Recommendations:

� It was the consensus opinion that lenalidomide as asingle agent has clinical activity in relapsed MCLand that LR was an option as initial therapy inuntreated MCL based on level B evidence.

� It was the consensus opinion that lenalidomide hasclinical activity in DLBCL based on level B evidence.

� The lenalidomide dose of 25 mg used in DLBCL ishigher than clinicians are accustomed to using inCLL; however, the risk of toxicity and clots/thrombosis is decreased for lymphoma patients.For patients without standard risk factors for deepvein thrombosis, the panel suggested givinglow-dose aspirin.

� The panel felt that clinical endpoints were needed tobetter define the duration of therapy for LR in FL.

Interferon (IFN)-α-2bIFN-α-2b belongs to family of interferons, which arenaturally occurring small proteins and glycoproteinsproduced and secreted by cells in response to viral infec-tions and to synthetic or biological inducers. Interferonsexert their effects through a complex sequence of


intracellular events including the induction of certainenzymes, suppression of cell proliferation, and augmen-tation of the specific cytotoxicity of lymphocytes fortarget cells [108]. IFN-α-2b is FDA-approved for the ini-tial treatment of clinically aggressive follicular NHL inconjunction with anthracycline-containing combinationchemotherapy in patients 18 years of age or older. Thisapproval was based on a randomized, controlled trialthat evaluated the safety and efficacy of IFN-α-2b inconjunction with a combination of cyclophosphamide,doxorubicin, and teniposide (CHVP) as initial treatmentin patients with clinically aggressive, large tumor burden,stage III/IV follicular NHL [109]. Patients were random-ized to CHVP alone or CHVP plus IFN-α-2b at 5 millionIU subcutaneously three times weekly for the durationof 18 months. The group receiving the combination ofIFN-α-2b plus CHVP had a significantly longer PFS (2.9years versus 1.5 years, P = 0.0001). After a medianfollow-up of 6.1 years, the median survival for patientstreated with CHVP alone was 5.5 years while mediansurvival for patients treated with CHVP plus IFN-α-2bhad not been reached (P = 0.004). IFN-α also has docu-mented single agent activity against multiple subtypes ofrelapsed NHL [110–112]. Direct injection of IFN-α intolymphoma lesions can often lead to their regression, sug-gesting that efficient delivery of IFN-α to tumors might bea useful approach to treating lymphomas [113, 114]. Toenable delivery of IFN-α to lymphoma cells, with anti-CD20 antibody-IFN-α fusion proteins have been devel-oped which show potent anti-lymphoma effects in pre-clinical models [115, 116].Recent evidence has also indicated that spontaneous ac-

tivation of the stimulator of IFN genes (STING) pathwaywithin tumor-resident DCs leads to type I IFN productionand adaptive immune responses against tumors [117].Lymphoma Panel Recommendations:

� The panel commented that IFN-α-2b is currentlynot commonly used in the treatment of NHL, andits indication came prior to the introduction ofrituximab. As such, its use should follow labelindications strictly or in the context of a clinicaltrial. However, other novel means of targetingIFN-α activities to tumor sites to treat lymphomasand other cancers are important areas ofinvestigation.

Emerging immunotherapies in lymphomaThere have been recent reports of several forms of im-munotherapy under clinical investigation for the treatmentof lymphoma that have demonstrated clinical efficacy. Asmany of these treatments are likely to receive FDAapproval in the coming years, the panel unanimouslyagreed that a brief overview of these modalities and clinical

data related to them would be of value to the prac-ticing oncologist. During the preparation of this manu-script, nivolumab received FDA approval for thetreatment of classical Hodgkin lymphoma (HL) thathas relapsed or progressed after autologous HSCT andpost-transplantation treatment with brentuximab vedo-tin. The subsequent section concerning checkpointblockade therapy was updated to reflect this approval.

VaccinesThere have been several trials evaluating the use ofvaccines in the treatment of lymphoma with one studyvalidating the vaccine approach by demonstrating im-provement of disease-free survival in a randomized, con-trolled clinical trial [118], while others have reportedvarying levels of success [119, 120]. As T cell activationis critical to a clinically relevant immune response, thereis a potentially a significant role for vaccines in the treat-ment of lymphoma, particularly in combination withother modalities. For vaccines to have a more significantrole, there is great need for new antigens, but unfortu-nately very few true tumor specific antigens in lymph-oma are known. Genome sequencing in context of HLAbinding permits the identification of large numbers ofneoantigens to which vaccines may be developed [121].The failure of vaccines may be due in large part to animmunosuppressive microenvironment, which may besecondary to past treatments or the inherent biology ofthe lymphoma. As such, there is a need to further under-stand vaccine efficacy in association with the microenvir-onment and develop biomarkers which will permit us toidentify subsets of patients or specific lymphomas thatmore likely to benefit from immunotherapy in general.

Cellular therapyThere are a variety of cellular therapies that have recentlydemonstrated clinical efficacy in lymphomas. These ther-apies include partially HLA-matched third-party Epstein-Barr virus (EBV)-specific cytotoxic T lymphocytes (CTLs),marrow and tumor-infiltrating lymphocytes (MIL/TIL),NK cells, and most prominently genetically-engineered Tcells, particularly CAR T cells targeting CD19 [122–125].

Third-party EBV-specific CTLs

� There is an increasing number of studiesdemonstrating that allogenic donor or “off-the-shelf ”third-party CTLs specific for EBV can be used safelyand successfully to treat EBV-associated lymphomas[122, 126].

� One donor can be used to generate antigen-specificT cells that can be infused into multiple recipientsmaking them readily and immediately available totreat patients.


CAR T cell therapy

� In contrast to the relatively large numbers andsuccesses in ALL and CLL, the use of CAR T celltherapy for the treatment of lymphoma is limited andhas short follow-up times. However, the available dataare encouraging with anecdotal data demonstratingresponses in refractory and relapsed FL, DLBCL, andMCL [125].

� There have been two major categories of toxicitiesassociated with this therapy: cytokine releasesyndrome (CRS) and neurologic toxicities, whichmay be related. Classical CRS is associated with highfever, tachycardia, hypotension, tachypnea andhypoxia, and it can be life-threatening [127]. CRS isassociated with elevated circulating levels of severalcytokines including IL-6 and IFN-γ, and uncon-trolled studies demonstrate that immunosuppressionusing tocilizumab, an anti-IL-6 receptor antibody,with or without corticosteroids can reverse thesyndrome. Neurologic toxicities observed withCAR-T cell therapy have included aphasia,dysphasia, tremor, and seizure. These havegenerally been transient, lasting up to 2 weeks,but they also can be life threatening.

� A significant practical obstacle in making thistechnology more broadly accessible is that thescreening and production process requires severalweeks. However, improving culture techniques havereduced production times to less than two weeks.There has also been increased standardization andautomation in manufacturing in preparation toprovide CAR T cells to large numbers of patients ascommercial products.

� A key scientific question for this field is why theresponse rates for lymphomas are so variable and notas high as those observed in ALL. One hypothesis isthat it may relate to host T cell function. A highlyrelated question is what is the optimal T cellphenotype for response and persistence, whichappears to correlate with duration of response [128].

� The majority of trials have targeted CD19, but CART cells targeting a number of other lymphomaantigens (e.g., CD22, CD28, CD30, ROR1) are inearly clinical trials or in development [129].

� This technology is very promising as a salvageregimen. However, the immediate question is its roleand timing among the many emerging choices forrefractory and relapsed lymphomas. There will beincreased utilization of this therapy and earlierconsideration for it as a treatment option, as long asit proves to be safe (see toxicities), and especially if itis shown to be a “once and done” option, which hasbeen observed in ALL.

Bispecific T cell Engager (BiTE) molecules

� Blinatumomab is FDA approved for the treatment ofrelapsed or refractory B cell precursor ALL. Itrecruits cytotoxic T cells to target tumor B cells bylinking the CD3 and CD19 antigens.

� In a phase II clinical trial, treatment of heavilypretreated patients with relapsed/refractory DLBCLwith blinatumomab showed an acceptable safetyprofile and resulted in objective (ORR = 43%) anddurable responses [130].

� CRS and neurotoxicity have been observed withblinatumomab.

Checkpoint blockadeTumor immune evasion pathways have been most thor-oughly studied in solid tumors; however, emerging datahave demonstrated that malignancies of hematopoieticorigin are also able to co-opt their local environment inorder to escape immune attack. Activated T cells upreg-ulate negative costimulatory receptors, such as PD-1 andcytotoxic lymphocyte antigen-4 (CTLA-4) [131]. En-gagement of PD-1 or CTLA-4 with ligands expressed ontumor cells or professional antigen presenting cells re-sults in down-regulation of effector T cell function andrepresents a potent mechanism of immune evasionacross a number of human cancers. Antibodies whichblock PD-1/PD-L1 interactions have demonstrated thatin select subtypes of HL and NHL, the PD-1 ligands areover-expressed due to a genetic amplification of the lociencoding them [132–134]. Other mechanisms of PD-L1over-expression in lymphomas have also been eluci-dated. Reports from early-phase clinical trials of PD-1blockade have demonstrated remarkable effectiveness inHL and also appear active against some NHLs.

� Preclinical studies suggested that Reed-Sternbergcells exploit the PD-1 pathway to evade immunedetection. In classic HL, alterations in chromosome9p24.1 increase the abundance of the PD-1 ligands,PD-L1 and PD-L2, and promote their inductionthrough Janus kinase (JAK)-signal transducer andactivator of transcription (STAT) signaling [133].Based on these observations, nivolumab, a PD-1-blocking antibody, was investigated in 23 patientswith relapsed or refractory HL [1]. An objectiveresponse was reported in 20 patients (87%) perinvestigator assessment, including 17% with a CRand 70% with a PR; the remaining 3 patients (13%)had stable disease. The rate of PFS at 24 weeks was86%. In a subsequent phase II study, nivolumab wasinvestigated in relapsed/refractory classical HLpatients. Results from this study illustrated anORR of 66% per independent review; CR and PR


rates were 8.8% and 57.5%, respectively. At thetime of the database lock for this study, 62% ofresponders remained in response with a medianfollow-up time of 8.9 months [135]. Based onresults from these studies, nivolumab was grantedaccelerated approval by the FDA on May 17, 2016for patients with classical HL that has progressedfollowing autologous HSCT and brentuximabvedotin.

� In trials with small numbers of patients, responseshave been observed with CTLA-4 or PD-L1blockade in FL and DLBCL [136–138].

� With virally-associated lymphoid tumors(e.g., EBV+ DLBCL), most all have increasedPD-L1 on tumor cells [132, 139]. Therefore,determining biological heterogeneity may allowfor the identification of subsets susceptible toPD-1 blockade.

� Trials of PD-1 blockade in lymphoma showtoxicities similar to those reported in solid tumors.

� Although results are very preliminary, the efficacyof PD-1 blockade as a single agent rivals that ofchemotherapy in heavily pretreated patients, andconsideration should be given to studying theseagents earlier in the disease course and incombination with conventional agents as wellas other forms of immune therapy, particularlyvaccines.

Issues related to immunotherapy research in lymphoma

� The panel thought it was essential to try to learn asmuch as possible from every patient who enters atrial. Specifically, it is important to obtain tumor andblood samples from every patient. Patient samplesare critical for evaluation of:

Tumor microenvironmentSystemic immune responsesTumor and host mutational burdenTumor antigensT cell receptor (TCR) repertoire (locally and

systemically) and clonal T cell expansion withintumors

� The panel suggested that pretreatment biopsiesshould be mandatory for participation in clinicaltrials and strongly suggested that follow-up biopsiesbe obtained at the time of relapse in order tounderstand mechanisms of resistance. In order todo so, there is a need for funding for sample banks.

� One of the major problems that will need to beaddressed is how to design and prioritizeimmunotherapy trials with so many competingagents and modalities. The panel suggested that aprofile/portfolio of collaborative immune studies

with uniform approaches to immune monitoring beestablished in order to develop a large dataset.

� It was emphasized that the majority of trials will bedeveloped and conducted with pharmaceuticalcompanies. Thus, it is imperative for industry toshare the biologic data that result from thesestudies. A collaborative effort is needed to bringtogether different interests and strengths in order todevelop important trial(s) and generate robust data.There is a strategic advantage to a pharma-academiapartnership. Such a partnership will result in fastercompletion of trials with greater scientific depth andwould be a “win-win” situation for both entities.

� In thinking about developing immunotherapeutictrials in lymphoma, the extraordinary heterogeneityof diseases, as well as within disease heterogeneity,must be recognized. Therefore, it is essential tostudy the quality and pathologic evidence ofimmune infiltration, which is the genetic basis forthe perturbation and modulation of regulators. Thisunderstanding of the biology and heterogeneitymust be linked to specific treatments for diseases

Acute leukemiaAcute myeloid leukemia (AML) and ALL remain for-midable clinical challenges largely due to resistance ofleukemia to current therapies and leukemia relapse[140, 141]. Negative immune regulatory mechanismspresent in acute leukemia may contribute to the devel-opment of a suppressive microenvironment that pro-tects leukemic cells from immune destruction.Furthermore, immune cell abnormalities includingimpaired NK cell activity and increased frequency andimmunosuppressive functions of regulatory T cellshave been described in patients with acute leukemia[142, 143].During the past four decades, allogeneic HSCT follow-

ing both myeloablative and non-myeloablative (reducedintensity) conditioning regimens has been established asa standard and curative treatment option for acuteleukemia [144–146]. The anti-leukemic activity of allo-geneic HSCT relies not only on the effects of high dosechemotherapy or irradiation given during the condition-ing regimen, but also on the immune-mediated graft-versus-leukemia effect [147–149]. The use of cytokinesor pharmacologic agents to restore immune cell effectorfunctions and, by extension, anti-leukemic effects repre-sent other immunotherapeutic approaches that havebeen used in leukemia treatment [150–153].Several non-transplant immunotherapeutic strategies

are currently being evaluated in numerous clinical trials.These include among others the use antibody basedtherapies, immune checkpoint inhibitors, CAR T cells,NK cells, and vaccine based therapies.


Current immunotherapies in acute leukemiaBlinatumomabBlinatumomab is a bispecific CD19-directed CD3 T cellengager that activates endogenous T cells when boundto the CD19-expressing target cell. Blinatumomab wasstudied in patients with MRD-positive B-lineage ALLafter intensive chemotherapy and in follow-up studies inpatients with relapsed and refractory Philadelphiachromosome-negative B cell ALL [154–157]. The role ofblinatumomab in is currently being evaluated in a PhaseIII clinical trial (ECOG-ACRIN Cancer Research Group,NCT02003222) in patients with newly diagnosed BCR-ABL-Negative B Lineage ALL.Blinatumomab was granted accelerated approval by

the FDA on December 3, 2014 for the treatment ofPhiladelphia chromosome-negative relapsed or refrac-tory B cell precursor ALL [155, 158]. The basis of theapproval was a single arm trial with 185 evaluable adults.Blinatumomab was administered in patients with refrac-tory/relapsed ALL by continuous infusion for 4 weeks ofa 6-week cycle. Up to two cycles were used for inductionand three cycles for consolidation. The complete remis-sion rate was 33% (95% CI: 27%–41%) with 2 cycles oftreatment with blinatumomab, and the median durationof response was 6.7 months (range, 0.46–16.5 months).Median OS was 6.1 months (95% CI: 4.2–7.5 months). Aminimal residual response was achieved by 31% (95% CI:25%–39%) of all patients.Safety was evaluated in 212 patients with relapsed or

refractory ALL treated with blinatumomab [158]. Themost common adverse reactions (≥20%) were pyrexia,headache, peripheral edema, febrile neutropenia, nausea,rash and tremor. Elevated transaminases were the mostcommon (>10%) laboratory abnormalities related to blina-tumomab. A neurological toxicity occurred in approxi-mately 50% of patients. CRS was reported in 12% of thepatients (grade 3 ≥CRS syndrome in 2%). Blinatumomabadministration was interrupted in 32% of the patients anddiscontinued in 17%. The most common reasons for

interruption were neurologic toxicity and CRS. Themost common reasons for permanent withdrawalincluded neurologic toxicity and sepsis.Leukemia Panel Recommendations:

� The panel recommended the use of blinatumomabfor patients with relapsed or refractory ALL basedon level B evidence.

Emerging therapiesMonoclonal antibodies in acute leukemiaEngagement of mAb with leukemia target antigens canlead to direct apoptosis, CDC, and ADCC [159]. Antigensexpressed on leukemia blasts or preferentially expressedon leukemia stem cells including CD33, CD45, CD96,CD123, CD135, CLL-1 and T cell immunoglobulinmucin-3 (TIM-3) represent potential targets for antibody-based therapy in AML [160, 161]. In ALL, CD19, CD20,CD22 and CD52 (among others) represent potential tar-gets [162–164]. A number of monoclonal antibodies arecurrently being evaluated (Table 1). These include uncon-jugated monoclonal antibodies and monoclonal antibodiesconjugated with cytotoxins.An approach to enhance the efficacy of antibody ther-

apy is the use of BiTE antibodies like blinatumomabmentioned on the previous page. By bridging tumorantigens with T cell receptors, these can direct effector Tcells to leukemia blasts target antigens. In recent years,different T cell engaging antibody constructs have beendeveloped. The use of bispecific antibodies that containCD16 and blast-specific antigens can enhance NK cellmediated ADCC. Furthermore, anti-KIR antibodies toblock inhibitory KIR receptors can be used to enhanceNK cell cytotoxicity [165, 166].Several phase I and phase II antigen-specific antibody

clinical trials are currently in development for the treat-ment of acute leukemia. Epratuzumab, an unconjugatedhumanized monoclonal antibody, binds to the third extra-cellular domain of CD22. Epratuzumab was evaluated by

Table 1 Selected monoclonal antibodies in ALL and AML

Selected monoclonal antibodies in ALL Selected monoclonal antibodies in AML

Rituximab anti-CD20 antibody SGN-CD33A anti-CD33 pyrrolobenzodiazepine dimer

Ofatumumab anti-CD20 antibody AMG 330 anti-CD33 and CD3, bi-specific T cellengager antibody

Epratuzumab anti-CD22 antibody MGD006 anti-CD123 and CD3, dual affinityretargeting molecule

Alemtuzumab anti-CD52 antibody CSL362 anti-CD123 antibody

Inotuzumabozogamicin

Monoclonal anti-CD22 immunotoxin SL-401 diphtheria toxin interleukin-3 fusion proteinagainst CD123

Blinatumomab bi-specific T cell engager antibody

Moxetunomabpasudotox

conjugated immunotoxin targeting CD22


the Children’s Oncology Group as single agent and as partof a chemotherapy backbone in 114 relapsed ALL patientseither weekly or twice weekly [167, 168]. The CR rateswere similar to both arms (65% and 66%) but were notsignificantly higher than those observed historically with-out epratuzumab. The addition of epratuzumab was welltolerated, with a similar toxicity profile to that observedwith the re-induction chemotherapy platform regimenalone. While CR rates were not improved compared tohistorical controls treated with chemotherapy alone, therewas a non-significant trend towards improvement inMRD response with the addition of epratuzumab to re-induction chemotherapy.In a recent SWOG study (31 patients, median age: 41

years old), the addition of epratuzumab to the combin-ation of clofarabine and cytarabine in adults with relapsed/refractory pre-B ALL was evaluated [169]. The responserate (CR plus CR without count recovery) was 52%, signifi-cantly higher than the previous trial with clofarabine/cytar-abine alone, where the response rate was 17%. The medianOS was 5 months (95% CI: 3–9 months).Rituximab, a chimeric anti-CD20 antibody, has been

evaluated with combination chemotherapy for patientswith B cell ALL demonstrating event-free survival (EFS)as well as OS benefit and molecular CR rates [170–172].A multicenter randomized trial compared a pediatric-inspired protocol to the same regimen plus rituximab inpatients newly diagnosed with CD20-positive Ph-negativeB-Cell precursor ALL (105 in the rituximab arm and 104in the control arm) [172]. Median age was 40 years. Bothrandomization arms were well balanced for pretreatmentcharacteristics. CR rate was 92% and 91% in rituximaband control arm, respectively. With a median follow-up of30 months, patients treated in the rituximab arm had alower cumulative incidence of relapse (CIR) (2-year CIR,18% [95% CI: 10–26] vs. 30.5% [95% CI: 21–40] in controlarm; P = 0.02), while no significant difference was ob-served regarding non-relapse mortality between botharms. This translated into longer EFS in patients treatedin the rituximab arm (2-year EFS, 65% [95% CI: 56–75] vs52% [95% CI: 43–63] in control arm; P = 0.038). Whencensoring patients who received allogeneic HSCT in firstCR at transplant time, EFS and OS were longer in therituximab arm.Ofatumumab is an anti-CD20 antibody that targets a

membrane proximal small-loop epitope on the CD20 mol-ecule. Similar to rituximab, ofatumumab was combinedwith ALL chemotherapy in a phase II clinical study. TheCR rate was 96%; and 96% of patients achieved MRDnegativity. The one year CR duration and OS were 90%and 88% respectively [173, 174].Alemtuzumab is a humanized monoclonal antibody

directed against the CD52 antigen present on the surfaceof immune cells. Alemtuzumab has limited activity as

single agent in patients with ALL [175]. In a phase Istudy by CALGB, alemtuzumab was administered post-remission for eradication of MRD. The addition of alem-tuzumab resulted in reduction of MRD, but it was alsoassociated with viral infections [176]. Based on theseresults an expansion phase was completed which mayconfirm the preliminary results.Inotuzumab ozogamicin is a humanized anti-CD22

antibody conjugated to calicheamicin. In a recent phase3 trial patients with relapsed or refractory ALL wererandomized to inotuzumab ozogamicin or standard ofcare intensive chemotherapy [177]. The rate of CR wassignificantly higher in the inotuzumab ozogamicingroup than in the standard-therapy group (80.7% vs.29.4%, P < 0.001). Among the patients who had CR ahigher percentage in the inotuzumab ozogamicin grouphad results below the threshold for minimal residualdisease. The duration of remission was longer in theinotuzumab ozogamicin group (median, 4.6 months vs.3.1 months; P = 0.03). In the survival analysis, which in-cluded all 326 patients, PFS was significantly longer inthe inotuzumab ozogamicin group (median, 5.0 monthsvs. 1.8 months; P < 0.001); the median OS was 7.7months (95% CI: 6.0 to 9.2) versus 6.7 months (95% CI:4.9 to 8.3), and the hazard ratio was 0.77 (97.5% CI,0.58 to 1.03) (P = 0.04). Inotuzumab ozogamicin hasalready received FDA Breakthrough Therapy Designa-tion for patients with relapsed or refractory ALL.CD33 is a myeloid differentiation antigen that is

broadly expressed on AML blasts. Antibody-based thera-peutics in AML have targeted CD33 for many years.Gemtuzumab ozogamicin is a targeted antineoplasticagent consisting of a recombinant anti-CD33 humanizedantibody linked to N-acetyl-γ-calicheamicin. Gemtuzu-mab ozogamicin was approved in 2000 by the FDA foruse in patients age 60 or older with CD33 + AML in firstrelapse [178, 179]. However, in 2010 gemtuzumab ozo-gamicin was voluntary withdrawn after a phase 3 trial(SWOG S0106) in newly diagnosed AML based showeda trend toward an increased mortality in the gemtuzu-mab ozogamicin arm [180]. A recent meta-analysis fromfive randomized controlled trials incorporating gemtuzu-mab ozogamicin demonstrated a significant survivalbenefit for patients with favorable and intermediate cyto-genetic characteristics suggesting of reassessing thestatus of gemtuzumab ozogamicin [181]. Given the po-tential of targeting CD33, new CD33 monoclonal anti-bodies are in development in clinical trials and CD33has been incorporated in bi-specific antibodies such asCD33/CD3 or CD33/CD123.SGN‑CD33, a CD33-directed antibody conjugated to

two molecules of a pyrrolobenzodiazepine dimer, hasbeen evaluated as monotherapy in patients with CD33-positive AML with CR + CRi rates up to 60% in


treatment naïve patients and in combination with hypo-methylating agents [182, 183].CSL362 is a fully humanized anti-CD123 monoclonal

antibody, engineered for greater ADCC by higher affinityfor NK cell CD16. An early report from a phase I clinicaltrial of 25 AML high-risk patients who achieved CR indi-cated that the antibody was safe and well tolerated [184].Leukemia Panel Recommendations:

� The panel recommended the use of rituximab inpatients with CD20-positive Ph-negative B-Cellprecursor ALL based on Level A evidence.

� All panelists agreed that mAbs should be evaluatedin clinical trials in the relapsed/refractory setting, innewly diagnosed acute leukemia patients withcombination chemotherapy, and in high-riskpatients in complete remission.

Immune checkpoint blockadeSurface expression and inhibitory functions of check-point inhibitors are up-regulated in T cells present in thetumor microenvironment. While the presence of theseinhibitory receptors on T cells is physiologically neces-sary to regulate cellular activation, their overexpressionin disease leads to dysfunction of T cells and other im-mune effector cells [185–187]. In the setting of cancer,chronic overexpression of checkpoint molecules resultsin T cell dysfunction and impairs anti-tumor immunity.The PD-1/PDL-1 pathway has been investigated in

preclinical leukemia mouse models. The PD-1 recep-tor biology, expression of PD-1 on the surface of acti-vated immune cells and its ligands, PD-L1 and PD-L2,on leukemic blasts and functional consequences ofantibody-based or pharmacologic blockade of PD-1are under investigation in acute leukemia [188–190].PD-1 blockade can restore anti-leukemia T cell func-tions and thus may offer therapeutic advantages inacute leukemia. Given the acceptable tolerability, pre-clinical rationale, and immunological activity of PD-1/PD-L1 blockade, clinical trials of anti-PD-1 mAbs areunderway in acute leukemia patients [191]. Several othercheckpoint molecules are known [192, 193] and are underinvestigation in acute leukemia, including CTLA-4, TIM-3, lymphocyte activation gene-3 (LAG-3), and B and T celllymphocyte attenuator (BTLA).Leukemia Panel Recommendations:

� The panel was in consensus that there is preclinicalrationale for consideration of clinical trials forimmune-checkpoint blockade in acute leukemia.

� The panel identified the following clinical settingsfor evaluation of immune-checkpoint blockade inacute leukemia: patients with MRD, high-riskpatients, and elderly patients.

CAR T Cells for the treatment of acute leukemiaAdoptive transfer of T cells engineered to express a CARhas emerged as a powerful immunotherapy. CAR-basedtherapies have been studied mainly in patients with Bcell ALL. As described above, CAR are synthetic mole-cules consisting of an extracellular antigen-bindingdomain fused via a spacer region to intracellular signal-ing domains that are capable of activating T cells. CARsengage molecular structures independent of antigenprocessing by the target cell and independent of MHC[194, 195]. Over the course of years, several generationsof CAR-T cells with different and multiple costimula-tory intracellular domains have been developed andtested in clinical trials [80]. First generation CAR includea single T cell stimulatory domain such as CD3-zeta.Second generation CAR add a co-stimulatory domainmost typically derived from CD28 or CD137 (4-1BB).Third generation CAR, not yet in clinical trials, include 2co-stimulatory signals. The later CAR generations withadditional intracellular signaling domains have increasedthe activity by circumventing the T cell’s need for co-stimulatory molecules. The addition of a co-stimulatorydomain in the new generation CARs improved the repli-cative capacity and persistence of modified T cells. Severalgene transfer technologies are used to engineer T cells toexpress CARs including electroporation as well as retro-viral and lentiviral vector methods.Most studies using CARs have focused on hematologic

malignancies by targeting CD19 [196]. Multiple clinicaltrials using other antigens are underway in ALL andAML. Reported clinical trials using CAR T cells differedin the design of the CAR, expression of the CAR on theT cells, conditions of the T cell culture, lymphodepletingstrategy, cytokine support for the infused T cells, andtiming of CAR T cell infusion with regard to standardtherapy such as allogeneic HSCT [80].High remission rates have been reported in patients

with relapsed/refractory ALL treated with CAR T cellswith CR rates of 70%-90%. Also, durable remissions wereobserved without additional therapy [80, 197–200]. Inaddition, in studies that included patients with priorhistory of allogeneic HSCT, no graft-versus-host diseasewas observed. Furthermore, among the different studies,the persistence of CAR-modified T cells varied, whichcould be related to different CAR design.Twenty-one children and young adults with ALL were

treated in a phase I clinical study with CD19-CAR in-corporating an anti-CD19 single-chain variable fragmentplus TCR zeta and CD28 signaling domains. Among 20patients with B-ALL, the CR rate was 70% (95% CI: 45.7–88.1), with 12 of 20 patients with B-ALL achieving MRD-negative complete response (60%; 95% CI: 36.1–80.9). OSat a median follow-up of 10 months was 51.6% at9.7 months and beyond. Leukemia-free survival of 12


patients who achieved an MRD-negative CR was 78.8%beginning at 4.8 months [201].Sixteen patients with relapsed or refractory B cell ALL

were treated in a phase I clinical study with autologous Tcells expressing the 19-28z CAR specific to the CD19 anti-gen. The overall CR rate was 88%, which allowed transi-tion of most of these patients to allo-HSCT. This therapywas as effective in high-risk patients with Philadelphiachromosome–positive (Ph+) disease as in those with re-lapsed disease after previous allogeneic HSCT [202].Thirty children and adults with relapsed or refractory

ALL were treated with autologous T cells transducedwith a CD19-directed CAR lentiviral vector that in-cluded the 4-1BB costimulatory signal (CTL019). CRwas achieved in 27 patients (90%), including 2 patientswith blinatumomab-refractory disease and 15 who hadundergone stem cell transplantation. At 6 months, theprobability that a patient would have persistence ofCTL019 was 68% (95% CI: 50 to 92), and the probabilitythat a patient would have relapse-free B cell aplasia was73% (95% CI: 57 to 94) [199]. Sustained remission wasachieved with a 6-month EFS rate of 67% (95% CI: 51–88%) and an OS rate of 78% (95% CI: 65–95%).CAR T cell therapies are associated with neurologic

toxicities including encephalopathy and seizures, whichcould be reversible. CRS is a common and potentiallylife-threatening toxicity associated with CAR T cell ther-apy [199, 201, 202]. It is associated with elevation ofcytokines including interleukin-2 receptor ɑ (IL-2Rɑ),IL-6, IL-10 and IFN.IL-6 inhibitors have been used to ameliorate CRS. The

development of CRS may correlate with response to ther-apy; patients who develop CRS often respond to therapy.Moreover, the severity of CRS may correlate with thetumor burden. Furthermore, B cell aplasia is an expectedon-target result of CD19-directed therapies and hasserved as a useful surrogate to determine the persistenceand effectiveness of CD19-directed CAR-T cells.The FDA has granted breakthrough therapy status to

CTL019, an investigational CAR therapy for the treatmentof pediatric and adult patients with relapsed/refractoryALL. CTL019 -engineered T cells express a CAR in whichthe T cell activation signal is provided by the CD3-zetadomain, and the costimulatory signal is provided by theCD137 (4-1BB) domain.Leukemia Panel Recommendations:

� The panel recognizes the lack of level A evidenceregarding the use of CAR T cells. However, all panelistsagreed that there are very encouraging clinical dataregarding the use of CAR T cells. With current-ongoing studies using CAR T cells, the panel concludedthat this treatment modality is among the most prom-ising immunotherapies for the treatment of ALL.

� The panel recommended the development ofguidelines for management of complications withCAR T cells.

Natural killer cells for the treatment of acute leukemiaNK cells contribute to anti-leukemic immune responsesboth by exerting direct cytotoxic activity and activatingother immune cells [203, 204]. In patients with acuteleukemia, there is impairment of NK cell activity medi-ated by several mechanisms including reduced expres-sion of activating receptors and reduced NK cellcytokine secretion capacity [205–208]. Furthermore, lowexpression of NK ligands on leukemia blasts and pro-duction of soluble immunosuppressive factors contributeto impaired NK cell activity [209, 210]. Different strat-egies for NK cell-based cancer immunotherapy havebeen studied including the administration of anti-KIRantibodies that can block NK cell inhibitory recognition,the use of cytokines including IL-15 and IL-2, adoptivelytransferred HLA-haploidentical NK cells and the use ofNK cell lines [203, 204, 211–213].In particular, adoptive therapy using haploidentical NK

cells in combination with cyclophosphamide and fludar-abine to lymphodeplete the recipient and facilitate ex-pansion of the allogeneic NK cells with interleukin (IL)-2 administration induced CR in patients with relapsedand refractory AML. Furthermore, the use of IL-2-diphtheria fusion protein (IL2DT) to deplete regulatoryT cells, led to improvements in rates of in vivo NK-cellexpansion and AML remissions (CR rate 53%) comparedwith patients that did not received IL2DT (CR rate 21%,P = 0.02) [212, 214].Leukemia Panel Recommendations:

� The panel recognizes the lack of level A evidenceregarding the use of NK cells. However, all panelistsagreed that there are very encouraging clinical dataregarding the use of NK cells.

� The panel identified the following clinical settings ashigh priority for evaluation of NK cell basedtherapies in acute leukemia: refractory patients,patients with MRD, and high-risk patients.

VaccinesTo stimulate anti-leukemic responses active immunizationsthrough vaccination has been explored for patients withacute leukemia. Several types of vaccines have been used forpatients with acute leukemia; peptide vaccines, granulocytesmacrophage-colony stimulating factor vaccines, and DCvaccines. Moreover, different maturation protocols havebeen applied for the generation of these vaccines. Pre-clinical and clinical trials involving vaccination with peptidesderived from different leukemia antigens, including Wilms’stumor 1 (WT1), PR1, or hyalunoric-acid-mediated motility


(RHAMM), demonstrated immunogenicity and clinicalsafety [215–221].Leukemia Panel Recommendations:

� The panel recognizes the lack of level A evidenceregarding the use of vaccines in acute leukemia.

� The panel agreed that clinical trials using vaccinesin acute leukemia have established their safety andimmunogenicity.

� The panel identified the following clinical settings ashigh priority for evaluation of vaccines in acuteleukemia: patients with MRD and high-risk patients.

� More studies are needed to investigate the effects ofchemotherapy on the immune system as intensivechemotherapy may impair the immune responses tovaccination.

Issues related to immunotherapy research in acuteleukemiaThe methods and timing needed to evaluate response toimmune therapies in acute leukemia are not established,and, similar to solid tumors treated with immune ther-apies, endpoints such as CR rates and PFS may need tobe re-evaluated.Leukemia Panel Recommendations:

� There are insufficient data to evaluate whethercurrent criteria for clinical response in acuteleukemia are adequate for evaluation of response toimmune therapies and whether immune relatedresponse criteria as in the setting of solid tumorswill be useful in acute leukemia.

� The panel recommended detailed immunemonitoring in ongoing clinical trials that are tailoredto specific immune therapies.

� The timing of the evaluation of response is not knownand should be tailored to the specific immune therapy.

� Standards of immune monitoring in acute leukemiashould be established and include among othersanalysis of bone marrow microenvironment andcirculating immune cells.

� Information gained from all on-going clinical trialsmay provide the foundation for the generation ofnew response criteria and the proper timing ofresponse evaluation.

Additional files

Additional file 1: Cancer Immunotherapy Guidelines for HematologicMalignancies Roster. (DOCX 13 kb)

Additional file 2: Comments Recieved during Review. (DOCX 14 kb)

Additional file 3: Annotated Bibliography (Myeloma), AnnotatedBibliography (Lymphoma), and Annotated Bibliography (Leukemia). (ZIP1812 kb)

AbbreviationsADCC: Antibody-dependent cell mediated cytotoxicity; ADCs: Antibody-drugconjugates; ALL: Acute lymphoblastic leukemia; AML: Acute myeloidleukemia; BiTE: Bispecific T cell engager; BTLA: B and T cell lymphocyteattenuator; CALGB: Cancer and leukemia group B; CAR: Chimeric antigenreceptor; CDC: Complement-dependent cytotoxicity;CHOP: Cyclophosphamide, adriamycin, vincristine, prednisone;CHVP: Cyclophosphamide, doxorubicin, and teniposide; CIR: Cumulativeincidence of relapse; CLL: Chronic lymphocytic leukemia; CR: Completeresponse; CRS: Cytokine release syndrome; CTL: Cytotoxic T lymphocyte;CTLA-4: Cytotoxic lymphocyte antigen-4; CVP: Cyclophosphamide, vincristine,and prednisone; CyBorD: Cyclophosphamide, bortezomib, anddexamethasone; DC: Dendritic cell; DLBCL: Diffuse large B cell lymphoma;EBV: Epstein-barr virus; EFS: Event-fr

Documents

The Society for Immunotherapy of Cancer consensus ... › content › pdf › 10.1186 › s40425-016-0188-z.pdfDavid Avigan6, Lisa Barbarotta7, Austin John Barrett8, Koen Van Besien9,